HBA's not sending alerts

[Morgan Leecy]

I am using a pair of DL380 G6's in a new SLES 10 OES cluster.

After a lot of pain, I have finally got the servers sending alerts (SNMP) to our SIM server, however the one error that is not being detected is when the HBA fails (or the fibre connection)

We are using 2x Qlogic 2x00 HBA's (set up with multipathing) and they function perfectly in all regards except this error reporting.

Can anyone advise on how to fix this issue. The servers are fully patched.

[Morgan Leecy]

The HBA's are HP Storageworks PCIe 8gb HBA's

There are no SMI-S CIM Providers for seperate download that I can see

[Rob Buxton]

The mechanism for alerts is that an agent picks up the issue and then gets SNMP to send the trap.
I would check that the FC Agents are being started. Not sure of the exact location but I think there's a config file (possibly cma.conf) that determines which sub-agentsa re started.
Also from memory there's a status parameter that can be sent to one of the hp programs that lists the agents running.
Sorry for being vague but we've mostly moved to ESXi and my experience was based on RedHat / ESX Consoles.